Last modified 3 years ago Last modified on 06/28/16 14:53:16

dCache Tier I meeting June 28, 2016

[part of a series of meetings]

Present, PIC(Marc), KIT(Samuel)


(see box on the other side)

Site reports


Everything running OK.

NFS and space-manager

Marc reminded us that they are still suffering from the NFS with space-manager issue.

AP/ Paul bring this up during the developers' meeting.

Marc described how they have implemented a work-around, involving by-passing the space-manager, and the users are looking at alternative protocols.

CMS readiness instance

Marc described how he plans to upgrade the CMS readiness instance to v2.16 in a couple of weeks. Due to pressure of work, he can't do this any earlier.


Samuel reported KIT has no issues currently.

PowerPC machines

Samual reported the PowerPC machines are not available any more.

Database deployment

Currently, KIT has two database instances per experiment: one for SRM (+...) and and one for chimera.

They want to have a single database instance for both, but run a master-slave deployment. The slave provides a hot stand-by for the master.

In this deployment, dCache would be configured to use a DNS alias that points to the master instance. If there is a failure, the DNS alias would be updated to point to the slave instance.

Would dCache need to be restarted after a DNS change?

AP/ Paul, ask dev. meeting.

Developing a NearlineProvider? plugin

KIT want to proceed in developing their own plugin, but are currently stuck. This is RT ticket 8983.

AP/ Paul, ask dev. meeting for progress on this.

Support tickets for discussion

[Items are added here automagically]


Same time, next week.